Local Analytical System for Early Epidemic Detection

Abstract

An Epidemic is a big threat to humanity. To reduce its catastrophic effect, many clinical practices and AI-based models are introduced to detect the onset of future Epidemic. An Analytical System can be useful for the prediction of an epidemic by collecting Quality data, modelling them and visualizing in different dimensions. This study deals with designing a Local Analytical System for early Epidemic detection in which the data related to human regular needs and responses are stored in the in-cube format. Analytical rules are used to produced faster pre-computed and pre-summarized inputs of the warehouse. Some desired inputs are selected from many local Warehouses which are then consolidated to form an incremental next higher-level data using the Layered Architecture style. This system can find the most commonly deviated data from the most frequently occurred patterns in the data submitted from the participating warehouses. The above-summarized patterns are mined using an FP-Growth algorithm to predict a new pattern. The patterns are ranked and inspected with their correlations for a possible unknown Epidemic.